[Rivet] rivet 1.3.0

[Sercan Sen]

Hi Andy,

please see my comments at the below.


On Oct 26, 2010, at 3:08 PM, Andy Buckley wrote:

> On 26/10/10 13:53, Sercan Sen wrote:
>>
>> Hi,
>>
>> I've just started to use rivet-1.3.0. My analysis is now throwing the
>> following error which I haven't had with the previous versions of  
>> rivet.
>>
>> Event 100000 (351 s elapsed)
>> Finished event loop
>> python: ../../include/LWH/Profile1D.h:362: virtual double
>> LWH::Profile1D::binError(int) const: Assertion `denom > 0' failed.
>> /var/spool/pbs/mom_priv/jobs/3134099.far.SC: line 9:  7830
>> Aborted                 rivet -a LHC_2010_EFWDDIFFGAP /home/ssen/
>> cedarssen/Data/P8-7TeV-ND.dat -H ND.aida
>>
>> But the problem is not only related with the analysis routine,  
>> because
>> sometimes it works fine if I give different events as input (I mean,
>> different HepMC file).  (maybe, it's looking for binError > 0 and if
>> it doesn't find it crashes...)
>
> Hi Sercan,
>
> This is a bug, yes: the assert throws when trying to calculate bin  
> error
> when there is (effectively) only one event in the bin. I've fixed this
> in the trunk version, so you might want to update to use that -- or
> alternatively patch the binError function in the include/LWH/ 
> Profile1D.h
> file: the line
>
> if (n_eff < 1.0) {
>
> should become
>
> if (n_eff <= 1.0) {
>
> Then rebuild and it should be ok. We'll produce a 1.3.1 patch fairly
> soon, and I just sneakily updated the 1.3.0 tarball to include the  
> patch
> since it's so small.
>
>> Also, although compare-histos/make-plots work fine (produces
>> meaningful histograms), rivet-mkhtml script does not produce a
>> meaningful output.
>
> Hmm, that's strange... it was working when I last tried. What happens?
> Anyone else seen this?


I run the standard MC_LEADINGJET analysis with the rivet version 1.2.1  
and 1.3.0, and put the "rivet-mkhtml" outputs to the following places.

http://www.hep.ucl.ac.uk/~ssen/DEBUG/plots121leadingjet/

http://www.hep.ucl.ac.uk/~ssen/DEBUG/plots130leadingjet/

And I realized that both of them are visible with Safari. However, I  
was using "konqueror" to look at the html outputs and the second one  
(1.3.0) did not work with konqueror.
Anyway this is good to learn, I can use firefox/safari...


>
> Thanks,
> Andy
>
> -- 
> Dr Andy Buckley
> SUPA Advanced Research Fellow
> Particle Physics Experiment Group, University of Edinburgh
>
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>